Real time multimedia workloads require progressingly more processing power. Modern many-core architectures provide enough processing power to satisfy the requirements of many real time multimedia workloads. When even they are un- able to satisfy processing power requirements, network-distribution can provide many workloads with even more computing power.
In this thesis, we present solutions that can be used to make it practical to use the processing power that networks of many-core architectures can provide. The research focus on solutions that can be included in our Parallel Processing Graphs (P2G) project.
We have developed the foundation for network distribution in P2G, and we have suggested a viable solution for execution of workloads on heterogeneous multi- core architectures.